Premium
Derivation of human health‐based ambient water quality criteria: A consideration of conservatism and protectiveness goals
Author(s) -
Tatum Vickie,
Wiegand Paul,
Stratton Steve,
Louch Jeffrey,
Ebert Ellen,
Anderson Paul
Publication year - 2015
Publication title -
integrated environmental assessment and management
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.665
H-Index - 57
eISSN - 1551-3793
pISSN - 1551-3777
DOI - 10.1002/ieam.1584
Subject(s) - percentile , human health , population , quality (philosophy) , probabilistic logic , environmental health , environmental science , risk analysis (engineering) , statistics , mathematics , business , medicine , philosophy , epistemology
Under the terms of the Clean Water Act, criteria for the protection of human health (Human Health Ambient Water Quality Criteria [HHWQC]) are traditionally derived using equations recommended by the US Environmental Protection Agency (USEPA) that include parameters for exposure assessment. To derive “adequately protective” HHWQC, USEPA proposes the use of default values for these parameters that are a combination of medians, means, and percentile estimates targeting the high end (90th percentile) of the general population. However, in practice, in nearly all cases, USEPA's recommended default assumptions represent upper percentiles. This article considers the adequacy of the exposure assessment component of USEPA‐recommended equations to yield criteria that are consistent with corresponding health protection targets established in USEPA recommendations or state policies, and concludes that conservative selections for exposure parameters can result in criteria that are substantially more protective than the health protection goals for HHWQC recommended by USEPA, due in large part to the compounding effect that occurs when multiple conservative factors are combined. This situation may be mitigated by thoughtful selection of exposure parameter values when using a deterministic approach, or by using a probabilistic approach based on data distributions for many of these parameters. Integr Environ Assess Manag 2015;11:298–305. © 2014 SETAC